1. Field of the Invention
The invention relates to a method for the control of display screens using the principle of modulation of the duration of light emission to display half-tones. It can be applied to screens where the pixels are constituted by cells working with two stable states and having a memory effect. The invention also relates to an image display device implementing the method.
The term "memory effect" is understood to mean the effect by which cells maintain either of two stable states when a signal having activated the state has disappeared.
2. Description of the Prior Art
Display screens of this kind are constituted for example by plasma panels (abbreviated as PP) of the direct current type with memory or the alternating current type, or again for example by screens whose elementary cells use a "point effect" phenomenon so that each one of them produces an electron beam.
In the case of alternating type PPs, the electrodes are covered with a dielectric material in such a way that, since they are not in contact with the gas, electrical charges collect on the dielectric at each discharge in the gas.
These charges persist at the end of the discharge and their presence in a cell then makes it possible to prompt a discharge in this cell by the application of a voltage lower than that which will be necessary in the absence of these charges. This constitutes the "memory effect" already referred to. The cells that possess such charges are said to be in the "lit" or ON state. The other cells which require a higher voltage to produce a discharge are said to be in the "extinguished" or OFF state.
This memory effect is used by means of alternating signals called sustaining signals, applied to all the cells to activate those that are in the "ON" state, namely to prompt so-called sustaining discharges in these cells that produce light without modifying their "ON" state or modifying the state of the cells that are in the "OFF" state.
All the so-called "alternating" PPs benefit from the above-described memory effect.
Certain alternating PPs use only two intersecting electrodes to define and activate a cell, as described for example in the French patent published under No. 2 417 848. In this case, the two intersecting electrodes are used to obtain both the addressing operation (namely the placing of the cell in the "ON" or the "OFF" state) and the sustaining discharges.
Reference may also be made to the "coplanar sustaining" type of alternating PPs, such as those known especially in the European patent document EP-A-0 135 382. In these PPs, each cell is defined at the intersection between a so-called addressing electrode and a pair of parallel electrodes. The sustaining discharges are carried out by means of the two parallel electrodes and the addressing is done by means of one of these two electrodes and the addressing electrode.
In the different types of PPs which show a memory effect, all the cells are supplied in parallel. The large number of cells possible may therefore lead to high currents and the supply of the cells may then show defects that generate defects of the image.
The authors of the invention have thought that the defects in the supply of the cells, due in particular to amplifiers working at the limits of their characteristics, are aggravated by the principle of the control of the half-tone of the image.
Indeed, the elementary cell of a PP has only two states: the "ON" state and the "OFF" state. Since it is not possible to have a similar modulation of the quantity of light emitted by a pixel, namely by a cell, the production of the half-tones is obtained by modulating the period of emission of light from the pixels in an image period, or in other words by modulating the time during which the cell is placed in the "ON" state within the image period.
The control and supply of the cells of a PP are explained here below.
FIG. 1 gives a schematic view of an alternating PP. To simplify the description, this PP is of the type with two intersecting electrodes to define a cell as described in the French patent No. 2 417 848 referred to further above.
The PP has an array of electrodes Y1 to Y4 called "row electrodes" intersecting with a second array of electrodes X1 to X4 called "column electrodes". Each intersection of row and column electrodes corresponds to a cell C1 to C16. These cells are thus arranged in rows L1 to L4 and in columns; CL1 to CL4.
Each row electrode Y1 to Y4 is connected to an output circuit SY1 to SY4 of a row control device 1 and each column electrode C1 to C4 is connected to an output circuit SX1 to SX4 of a column control device 2.
The working of these two control devices 1, 2 is managed by an image management device 3.
Each output SY1 to SY4 of the row control device 1 delivers voltage square-wave signals that form the above-mentioned sustaining signals. These sustaining signals are thus applied simultaneously to all the row electrodes Y1 to Y4.
FIGS. 2a to 2d show sustaining signals applied respectively to the row electrodes Y1 to Y4. FIG. 2a particularly shows that the sustaining signals are formed by a succession of voltage square-wave signals set up on either side of a reference potential Vo which is often the potential of the ground. These square-wave signals vary between a negative potential V1 where they show one plateau and a positive potential V2 where they show another plateau. The reference potential Vo is applied to the column electrodes X1 to X4 in such a way that the application of the sustaining signals develops alternately positive and negative voltages at the terminals of the cells C1 to C16, for example voltages of 150 V that generate discharge in all the cells of the PP that are in the "ON" state.
These discharges occur at each reversal of polarity of the sustaining square-wave signals, namely at each positive transition Tp and negative transition Tn of these signals.
The placing of the cells in the "ON" or "OFF" states is done by addressing operations that are managed by the image management device 3. They may consist for example of the superimposition of specific signals of the addressing operation on the square waves of the sustaining signals. To this end, the row electrodes Y1 to Y4 are individualized, namely they are connected to an output circuit SY1 to SY4 proper to each of them, and each output circuit has, for example, a mixing circuit (not shown) by means of which it receives the sustaining signals and the addressing signals which come from different channels.
The sustaining signals have a period P that may be for example 10 microseconds, during which there address all the cells belonging to a selected row L1 to L4, namely all the cells defined by means of a selected row electrode Y1 to Y4.
Assuming that, at an instant to, there starts the addressing of the first row L1 corresponding to the row electrode Y1, the addressing may be for example of a type such that, at this instant to, the signal applied to this electrode Y1 (and only to this electrode) is a negative erasure transition Tne, with a duration (shown in dashes) greater than that of the other transitions, that causes all the cells connected to this row electrode Y1 to be placed in the &lt;&lt;OFF&gt;&gt; state. Then, at an instant t1 when the signal has its positive plateau, a so-called recording square-wave signal C1 (shown in dashes) is superimposed (on the positive side) on this plateau. This recording square-wave signal has the effect of placing all the cells connected to this row electrode in the "ON" state except those whose column electrodes X1 to X4 deliver a so-called "masking" signal (not shown) which has the effect of inhibiting the effects of the recording square-wave signal CI.
This operation may be repeated at each of the following periods of the sustaining signals at the instants t2 and t3, t4 and t5, t6 and t7 at which the operations for addressing the rows L2, L3, L4 corresponding respectively to the row electrodes Y2, Y3, Y4 are thus made. At the instant t8, a new addressing of the first row L1 is carried out.
These addressing operations, performed successively for each row L1 to L4 of the screen, constitute a sub-scanning operation, and several sub-scanning operations are performed during an image cycle time or image period in order to obtain the half-tones of the image by placing the cells C1 to C16 of each row L1 to L4 in the "ON" state or the "OFF" state at each sub-scanning operation.
To this end, the image period PI is sub-divided into n sub-periods S1, S2, . . . , Sn of different duration, respectively equal to To, 2To, . . . , 2.sup.n -1 To, with To=PI/2.sup.n -1.
FIG. 3 illustrates the division of the image period PI into n sub-periods S1, S2, . . . , Sn with n equal to 4 in the example. The image period PI starts at the instant to with a first sub-period S1 that lasts for a period of time To and ends at an instant Ta. A second sub-period S2 starts at the instant ta and lasts for a period of time equal to 2To, ending at an instant tb at which a third sub-period S3 starts. The third sub-period S3 lasts for a period of time equal to 4To and ends at an instant tc. A fourth sub-period S4 starts at the instant tc and lasts for a period of time equal to 8To up to the end of the period PI which marks the instant to' of a following image period.
With an image period PI equal to 20 ms for example, the sub-periods S1, S2, S3, S4 respectively have a duration of the order of 1.33 ms, 2.66 ms, 5.33 ms and 10.66 ms.
Thus, in the example of FIG. 3, it is possible to address each row L1 to L4 four times during the image period of this row, at the instants to, ta, tb and tc. It is therefore possible, for each row L1 to L4, to place each cell C1 to C16 in the "OFF" state or the "ON" state at each of these instants, namely at each start of the sub-periods S1 to Sn, and each cell preserves this state up to the beginning of the next sub-period when it is again placed in either of the two states, namely the "OFF" state or the "ON" state.
The cells that have been placed in the "ON" state by the beginning of one or more sub-periods S1 to Sn are activated by the sustaining signals and produce light for the duration of this sub-period or these sub-periods. It is therefore possible, by the combination of the n sub-periods S1 to Sn, to obtain 2.sup.n -1 different periods of emission of light by each cell. Each period corresponds to a desired luminance level for this cell during the image period PI. In addition, there is a period corresponding to the zero luminance level which corresponds to the case of a cell that is placed in the "OFF" state for all the periods S1 to Sn of this image period.
Thus, in the example of FIG. 3, the luminance level of a cell placed in the "ON" state, namely a cell activated solely during the first sub-period S1, is 1/5th of the luminance level activated during the first and third sub-periods S1, S3 and 1/15th of the luminance level activated during the entire image period PI.
This principle of the control of the luminance levels of the cells of a row L1 to L4 can be applied to all the rows, of course with a time lag from one row to another. For example, the principle can be applied from a row L1 to the following row L2 with a lag that corresponds to a sustaining signal period p as shown in FIG. 2 which may, for example, be in the range of 10 microseconds. In fact, the image period PI has one and the same duration for all the rows L1 to L4, irrespective of the number N of these rows with a time lag for example of one period between two consecutive rows. This lag is seen again in the sub-division of the sub-periods S1 to Sn.
It must be noted that the luminance levels desired for the different cells of each row L1 to L4 correspond to video input luminance values that are encoded and stored in the image management device 3, generally by means of n bits of different values of significance, each corresponding to one of the sub-periods S1 to Sn.
Since the cells C1 to C16 in the "ON" state are activated by the sustaining signals delivered by the row control device 1, they constitute a load applied to this device.
The sustaining signals may be prepared in different ways that are known per se. In any case, the row control device has at least one amplifier A for this purpose. This amplifier A delivers the sustaining signals to the output circuits SY1 to SY4 either directly as shown in FIG. 1 or by means of several output stages (not shown) each assigned to the supply of several output circuits, namely several row electrodes Y1 to Y4.
In the example of FIG. 1, only four row electrodes and four column electrodes are shown but a PP may actually have more than a thousand electrodes of each of these types, which define more than a million cells.
Consequently, the sustaining signals delivered by the amplifier A must be delivered by the amplifier at a current that may vary considerably as a function of the contents of the image, namely as a function of the number of cells that are in the "ON" state. Given the non-zero source impedance values of the amplifier A, as well as the impedance values of access to the cells (related in particular to the inductance and resistance values of the printed circuit tracks and the connections etc.), the quantity of charges really applied to a given cell C1 to C16 depends on the total content of the image. In other words, the greater the load applied to the amplifier A, the greater is the reduction in the luminance of the &lt;&lt;ON&gt;&gt; cells forming this load.
This variation of luminance as a function of the contents of the image can be seen especially in the case shown in FIG. 4 which represents an image that is formed chiefly by a low-luminance peripheral zone Z1 and a second high-luminance zone Z2 and has a constant encoding of video luminance value. A major variation of luminance displayed can be observed as a function of the variation of the surface area of the second zone Z2.
To this defect of the image, there is added another defect related to an excess brightness defect. This defect consists especially of an exaggeration and even a reversal of the differences in luminance between zones. Here, reference is made again to FIG. 4, assuming that the second zone Z2 is formed by two contiguous surfaces R1, R2, the second one R2 of which is located at the center of the first one R1 and assuming that it is desired to display luminance values that are different but close to each other on these two surfaces, for example a luminance I2 corresponding to a video luminance encoding equal to 128 (in the case of a video luminance encoding on 8 bits, namely with eight sub-periods as explained here above), for the second surface R2 and a luminance I1 encoded 127 for the first surface R1.
An exaggeration and a reversal of the difference in luminance displayed by these two surfaces R1, R2 can be observed in the image: instead of a theoretical ratio I2/I1 of 1.008 (128/127=1.008), in fact a real ratio is obtained with a value that may be 0.54.
Furthermore, if the first surface R1 is made to vary, all else being equal, a variation is noted in the luminance values I1 and I2. This variation shows that the luminance I2 of the second surface R2 is dependent on the contents of the rest of the image outside this surface R2.
A known approach used to correct these defects consists in diminishing the impedance values of sources and the impedance values of connections, and the impedance values presented by the electrodes themselves. This is obtained by a choice and selection of the components and by drawing and preparing the paths of the discharging currents with special care and also by increasing the number of channels provided for the discharge currents (especially by the parallel connection of several power transistors at the sustaining signal amplifier or amplifiers, such as the amplifier A, as well as in the output circuits, such as the circuits SY1 to SY4).
However, the improvements that result therefrom are only partial and, at the same time, very costly for the number of components and/or their individual cost is greatly increased. There is also an increase in both space requirement and manufacturing complexity, and hence an increase in cost.